The Investigation of DNA-Methyltransferase Interactions in the Adenine Methyltransferases using the Time-resolved Fluorescence of 2-Aminopurine
Bonnist, Eleanor Y M
The time-resolved fluorescence of 2-aminopurine (2AP) has been used to investigate DNA base flipping by the adenine methyltransferases and to study aspects of the DNA-enzyme interaction. 2AP is an excellent fluorophore to probe base flipping in the adenine methyltransferases because, as demonstrated in the present work on M.TaqI, the 2AP is delivered into the same position inside the enzyme as the natural target adenine and with the same orientation that prepares the adenine for enzyme catalysis. 2AP emits two types of fluorescence when in DNA. The first is the well-known 370-nm emission, which emanates from 2AP as a monomer species. The second is 450-nm emission and comes from a 2AP which is π-stacked with a neighbouring DNA base, a heterodimer species. Additionally, 450-nm emission is produced by a 2AP-tyrosine or 2APphenylalanine heterodimer when a flipped 2AP is π-stacked inside a DNA-methyltransferase complex. Steady state fluorescence of the 2AP-heterodimer has been used to complement the time-resolved investigations. Combined crystal- and solution-phase studies on M.TaqI have shown that when 2AP is flipped into the active site of M.TaqI it is significantly quenched by face-to-face π- stacking with the tyrosine from the NPPY catalytic motif. Not all of the flipped bases are held inside NPPY; in a minority of complexes, the flipped 2AP experiences very little quenching within the interior of the enzyme. In the sequence of bases recognised by M.TaqI, the thymine opposite the target adenine does not actively cause base flipping, as previously suggested, however, its presence aids the successful delivery of the target base into NPPY. For the DNA-M.TaqI-cofactor ternary complex, the effect of varying the cofactor has been investigated. The use of 5’-[2(amino)ethylthio]-5’-deoxyadenosine (AETA) or sinefungin as cofactor analogue causes M.TaqI to show different base flipping behaviour compared with the natural cofactor S-adenosyl-L-methionine (SAM) and with the cofactor product S-adenosyl homocysteine (SAH). In the ternary complex containing SAM the flipped base is held the most tightly within the catalytic motif. M.TaqI mutants have been studied in which the tyrosine (Y) in the NPPY motif is mutated to alanine (A) or phenylalanine (F). Stabilisation of the flipped base inside these mutants is more reliant on edge-to-face π-stacking with phenylalanine 196 and the available hydrogen-bonding in the adenine binding pocket. The NPPF-phenylalanine does not π-stack with the flipped base as NPPY-tyrosine does. Solution-phase time-resolved fluorescence studies have confirmed that M.EcoRI and M.EcoRV use a base-flipping mechanism to extrude their target bases. For M.EcoRI, with sinefungin cofactor, the majority of the flipped 2APs are not held in the NPPF catalytic motif. When the natural SAM cofactor is used, however, the flipped 2AP strongly associates with NPPF inside M.EcoRI. Non-cognate sequence binding has been investigated, in which M.EcoRI encounters a base that is in almost the same sequence context as the methylation target. M.EcoRI forms some direct contacts with the pseudo-target adenine but does not extrude the base that is in a highly stacked position inside the duplex. The H235N mutant of M.EcoRI, measured under the same conditions as the wild-type enzyme, shows different behaviour to the wild-type enzyme in a small proportion of complexes, when bound to the cognate recognition sequence, and is far more discriminating than the wild-type when bound to the non-cognate sequence. The M.EcoRV methyltransferase was found to be less efficient at flipping its target base than with M.TaqI or M.EcoRI. When M.EcoRV binds to its GATATC recognition sequence, the base-enzyme interactions of the target (GAT) and non-target (TAT) adenine position are shown to be quite different.